Using epidemiologic techniques, we will: 1. investigate the role of genetic factors and modes of genetic transmission in subtypes of depressive disorders by segregation and linkage studies of pedigrees; 2. develop more precise genetic models and apply these new as well as existing models to the data; 3. develop more precise clinical definitions of the familial manifestations of mood disorders; 4. explore the relationship between a variety of social and clinical characteristics of probands and morbid risk of psychiatric disorder in their relatives. Approximately 2,400 subjects will be interviewed directly (400 probands and 2,000 of their first-degree relatives). The probands will be selected from four distinct populations and will include a broad spectrum of affective illness (bipolar, severe unipolar, less severe unipolar depressives) and a matched nonpsychiatric normal control group. Approximately four first-degree relatives and a spouse will be interviewed per proband, including the relatives of the normals. Additional information will be collected by family history. A population survey of the life-time prevalence of psychiatric illness has been conducted by the investigatos. The population data will provide an additional comparison. Information for determining the diagnosis of relatives and probands will be obtained by detailed structured interview (Schedule for Affective Disorders and Schizophrenia) and will be applied to precise diagnostic criteria (Research Diagnostic Criteria). Identical methodology and diagnostic criteria will be applied to all probands, relatives, and population subjects.